Asymmetric Membranes for Gas Separations - American Chemical

11 Asymmetric Membranes for Gas Separations

Downloaded by UNIV OF PITTSBURGH on December 23, 2014 | http://pubs.acs.org Publication Date: December 12, 1985 | doi: 10.1021/bk-1985-0269.ch011

HEINZ FINKEN

1

Department of Chemical Engineering, The University of Texas at Austin, Austin, TX 78712

Recent membrane developments for gaseous mixture separations are compared to the development of reverse osmosis membranes for water desalination. The goals of these developments have been the search for ideal permselective polymeric materials, techniques for producing ultrathin membrane layers free of imperfections and transforming gelled reverse osmosis membranes into solid gas permeation membranes. A novel approach to meeting the basic requirements of high permselectivity is attempted by altering the physical polymer structure within the membrane prior to application for gas separation. The influence of these physical interactions on membrane properties is presented. The development of membranes f o r gas separations has been closely linked to that f o r water desalination. Therefore, a short review of the history of reverse osmosis membranes i s f i r s t given, followed by a presentation of the milestones leading to today's commercial, asymmetric membranes f o r gas separations. Membrane Development f o r Water Desalination The era of economically v i a b l e membrane development which began i n the l a t e 1950's and continues to t h i s date, may be divided into two time periods: the f i r s t generation of integral-asymmetric, c e l l u l o s i c membranes (1959 to 1970) and the second generation of asymmetric, non-cellulosic membranes (1971 to 1984).

Current address: Institut fuer Chemie, GKSS Forschungszentrum, Geesthacht GmbH, Max-Planck-Strasse, D-2504 Geesthacht, Federal Republic of Germany

0097-6156/85/0269-0245$08.00/0 © 1985 American Chemical Society

In Materials Science of Synthetic Membranes; Lloyd, D.; ACS Symposium Series; American Chemical Society: Washington, DC, 1985.


246

M A T E R I A L S SCIENCE O F SYNTHETIC M E M B R A N E S

C e l l u l o s i c Membranes. In the f i r s t period, summarized i n Table I, c e l l u l o s e acetate (CA) was discovered as a highly selective material by Reid and co-workers (1) who found high sa^t rejections, but unfortunately low permeate water fluxes (40 L/m d) through t h e i r membranes. In 1960, Loeb obtained widely scattered results during permeation measurements of CA u l t r a f i l t r a t i o n membranes manufactured by the German company Schleicher & Schuell. This was f i r s t attributed to sealing problems, but l a t e r found to depend on the side of the membrane which was exposed to the feed water (2). He interpreted t h i s behavior to be due to the asymmetric structure of the membranes, which was confirmed l a t e r by scanning electron miscroscopy. Loeb and Sourirajan (3) were successful i n developing a procedure f o r preparing asymmetric membranes. They invented a solvent-casting/water-precipitating method. The membranes consisted of a 0.1 to 1.0 micron-thin skin layer and an integrally-bound, 100 to 200 micron-thick supporting layer. The skin layer was non-porous and so t h i n that dissolved s a l t s were rejected almost completely, and water was allowed to permeate through i t with £luxes 10 times greater than those of symmetric membranes (400 L/m d). The spongel i k e , microporous sublayer had mechanical s t a b i l i t y which allowed for high pressure operation and produced high permeate fluxes and recoveries. The decade following t h i s invention was devoted to improvements of the transport properties and s i m p l i f i c a t i o n s of the preparation procedures. As CA membranes d i d not meet the s a l t r e j e c t i o n requirements necessary f o r one-stage seawater desalination, a search for more favorable materials started. In 1965, Merten and co-workers (4) prepared a double-layer CA membrane by solventcasting on a water surface and laminating the u l t r a t h i n f i l m on a microporous supporting membrane. A s a l t r e j e c t i o n f o r seawater of 99.8% was achieved. This type of membrane i s now c a l l e d a " t h i n f i l m composite membrane." The same preparation technique was used some years l a t e r by Cadotte and co-workers (5) to make 500 X thin layers of c e l l u l o s e acetate of varying acetyl content. A c e l l u l o s e diacetate membrane with an acetyL content of 39.8% yielded a permeate flux of more than 650 L/m d with a s a l t r e j e c t i o n of 94%, whereas a c e l l u l o s e t r i a c e t a t e membrane with an acetyl content of 43.2% exhibited a lower permeate flux of 200 L/m d at a higher s a l t r e j e c t i o n of 99%. A l l these membranes were only suitable f o r onestage brackish-water desalination. A breakthrough f o r seawater desalination occurred i n 1970 with the findings of Cannon, S a l t o n s t a l l and co-workers (6,7) that a blend of c e l l u l o s e diacetate and triacetate resulted i n membranes showing better permselectivities than those prepared from the single components. Blend membranes with an average acetyl content of 41.5% gave i n i t i a l permeate water fluxes of 530 L/m d and s a l t r e j e c t i o n of 99.6% when tested with 3.5% NaCl-solutes at 25°C and 105 bar. Non-cellulosic Membranes. Despite an intensive search f o r more favorable membrane polymers, c e l l u l o s e acetate remained the best material f o r reverse osmosis u n t i l 1969 when the f i r s t B-9 permeator for brackish water desalination was introduced by Du Pont. Richter and Hoehn (8) invented aromatic polyamide asymmetric hollow-fiber


g

Ρ

£

Ο

integral-asymmetric CA membranes

Loeb and

1960

1977

1969

ΣΗ.1970

- » —τ Ct> « Û) =3

integral-asymmetric CA blend

t h i n - f i l m composite CTA membranes

of CA

(9,10)

Cadotte et a l .

Hoehn (8)

Richter and

membranes

interfacial-polymerization

TFC

aromatic polyamide hollow-fibers

S a l t o n s t a l l (6,7) membranes

Cannon and

(D

Cadotte et a l .

(A)

Merten et a l .

t h i n - f i l m composite (TFC) membranes

selection of c e l l u l o s e acetate (CA)

Reid et a l . (1)

1959

Sourirajan (3)

Invention

Inventor

desalination

(98.6%) (99.86%)

(99.6%)

ο

high s a l t r e j e c t i o n

(99.5%)

extremely high water flux (800 L/m d) at

rejection

low-cost module system with high s a l t

salt rejection

higher water flux (500 L/m d) at high

flux (200 L/πι d)

high s a l t r e j e c t i o n (99%) at medium water

high s a l t r e j e c t i o n

rejection

increased water f l u x (400 L/m d) and s a i t

r e j e c t i o n of 96%

low water flux (40 L/m d) at a s a l t

Main Improvement

Table I: Milestones In membrane development f o r water


Year

ET. O1970

g; CD

Ο

36



248


membranes f o r t h i s module. Although t h i s moderately hydrophilic, rather s t i f f - c h a i n ^ polymer showed an i n t r i n s i c a l l y lower water permeation (70 L/m d) than c e l l u l o s e acetate, i t became competitive because of the low-cost self-supporting f i b e r technology and i t s high packing density. £ hollow f i b e r module can reach packing densities up to 9200 m /m compared to a spiral^wound module equipped with f l a t sheet membranes with only 660 m /m . The comparison of these systems i s more r e a l i s t i c on the basis of t h e i r water flux densities:2 a va^lue of 680 m /d/m i s calculated for a hollowf i b e r and 230 m /d/m f o r a spiral-wound module system. In 1977, Cadotte and co-workers (9) combined the technique of t h i n - f i l m composites with i n t e r f a c i a l polymerization. This new method extended the range of possible membrane polymers to watersoluble compounds l i k e polyethyleneimine which were s t a b i l i z e d after casting by cross-linking with, f o r example, toluene diisocyanate. The i n t r i n s i c a l l y high water permeability of these hydrophilic polymers along with the technique of making 500 & t h i n , selective layers free of imperfections revealed a membrane with a permeate flux of 800 L/m d at a s a l t rejection of 99.5% (North Star, NS 100). Aside from these decisive milestones i n the development of reverse osmosis membranes, further advances have been achieved which are important i n terms of r e p r o d u c i b i l i t y , a v a i l a b i l i t y , flux improvements (more than 1100 L/m d at 99.5% have been achieved to date), mechanical s t a b i l i t y and chemical resistance. The inventions are l i s t e d here i n chronological order. Their appropriate treatment would require a s p e c i a l discussion outside the realm of t h i s paper: 1966 - CTA hollow-fiber system (Dow) (10); 1975 - f l u x - s t a b i l i z e d CA blend membranes (GKSS, M 97 TVZ) Qi); 1976 - t h i n - f i l m composite membranes of cross-linked polyetheramide (UOP, PA 300) (12)5 1977 - polybenzimidazolone, t h i n - f i l m composite membranes ( T e i j i n ) (13); - t h i n - f i l m composite membranes of cross-linked polyether/urea (UOP, RC 100) (14); 1980 - t h i n - f i l m composite membranes of cross-linked polyether (Toray, PEC 1000) (15); - t h i n - f i l m composite membranes of modified, cross-linked aromatic polyamide (Film Tec, FT 30) (16). The Development of Asymmetric Membranes f o r Gas Separations C e l l u l o s i c Membranes. The f i r s t asymmetric membrane f o r gas separation appeared i n 1970 (Table I I ) , and i t was not surprising that t h i s membrane was a modified CA membrane of the Loeb-Sourirajan type (17). Gelled CA membranes for water desalination must be stored wet i n order to maintain t h e i r permeation performance. However, i n gas permeation, wet, p l a s t i c i z e d membranes tend to lose t h e i r properties with time due to p l a s t i c creep of the soft material under pressure and due to slow drying during which the microporous sublayer may collapse and thus increase the thickness of the dense skin-layer. Gantzel and Merten (17) dried CA membranes with an acetyl-content of 39.4% by quick-freezing and vacuum sublimation at


In Materials Science of Synthetic Membranes; Lloyd, D.; ACS Symposium Series; American Chemical Society: Washington, DC, 1985. at a He/N

extremely high N -2 Π

membranes

TFC membranes made of s i l i c o n e /

asymmetric polysulfone hollow

f i b e r s coated with s i l i c o n e

Ward et a l . (31)

Henis and

T r i p o d i (34)

1976

1979

2

2

2

^

2

2

selectivity

permeation rate of ' at a low 0 /N

0

s e l e c t i v i t y of 68

permeation rate of 4.8x10 ^ ^

s e l e c t i v i t y of 34

6

cost module system

high H /C0 s e l e c t i v i t y of 33 with a low-

of 2.3

8.6x10

1) P/ Jp an (STP)/cm -s-cmHg; to obtain m (STP)/m -s-MPa multiply by 7.501.

rubber

polycarbonate block copolymers

higher

solvent-dried, asymmetric CA blend

S c h e l l (20)

1975

at a He/N

permeation rate of 3.1xl0~

membranes

2

Merten (17)

high N

freeze-dried, asymmetric CA

Gantζel and

Main Improvement

1970

Invention

Inventor

Milestones i n membrane development f o r gas separation

Year

Table I I :


M A T E R I A L S SCIENCE OF SYNTHETIC M E M B R A N E S

250

-10°C. They obtained nitrogen permeation rates (permeability, P, per unit thickness, Z) of 3.1 χ 10" cm (STP)/cm -s-cmHg with a variable-volume/constant-pressure method at 22°C and a separation factor for He/N of 34. The higher separation factor of 97 observed with a f u l l y dense, symmetric membrane sample of i d e n t i c a l chemical composition was explained by the presence of defects i n the skinlayer of the asymmetric membrane. Four years l a t e r , Stern and co-workers (18) investigated commercial desalination CA membranes with an acetyl-content of 40.8%. The membranes were dried by a simple method developed by Vos and Burris (19): f i r s t soaking them i n an aqueous solution containing surfactants to reduce the water-polymer surface tensions, and then air-drying thenu The nitrogen permeation rate of such a membrane was 4.7 χ 10" cm (STP)/cm -s-cmHg at 30°C measured with the time-lag method which i s a constant-volume/ variable-pressure method with a downstream pressure of almost zero. Although t h i s value indicates lower transport properties, the authors could demonstrate that the permeation of permanent gases through CA membranes can be described accurately by a solution/ d i f f u s i o n mechanism. The same idea of drying membranes with solutions of decreasing surface tension was applied by S c h e l l (20) to integral-asymmetric CA blend membranes. The membranes were prepared through a sequence of successive solvent exchange steps followed by a i r - d r y i n g . A t y p i c a l permeation rate for nitrogen i s 4.8 χ 10" cm (STP)/ cm -s-cmHg at 25°C. This value i s better than that observed for CA membranes by Gantzel and Merten (17). The He/N^ separation factor of 68 i s twice that observed f o r a CA membrane. However, the higher i n t r i n s i c value of 97 for a dense, symmetric c e l l u l o s e diacetate f i l m could not be reached. Nevertheless, a high separation factor of 68 for an asymmetric CA blend membrane proves that i t s skin-layer must be dense and free of imperfections. The decreased hydroxyl content of the CA blend compared with the c e l l u l o s e diacetate homopolymer diminishes the intermolecular hydrogen bonding of the polymer chains and leads to an increase of polymer segments mobility and to an easing of d i f f u s i o n of penetrating gas molecules. These considerations explain the higher permeabilities for both helium and nitrogen gases through CA blend membranes at the expense of lowered s e l e c t i v i t y . CA blend membranes f o r gas separation are commercially available from Envirogenics Systems Co. ( E l Monte, CA), Separex Corp. (Anaheim, CA) and Grace Membrane Systems (Houston, TX), and are applied i n spiral-wound modules for the separations of a c i d i c gaseous components from natural gas, f o r the recovery of carbon dioxide i n enhanced o i l recovery processes for gas dehydration or the separation of hydrogen from carbon monoxide (21-23).


2

Non-cellulosic Membranes. While the development of CA gas permeation membranes can be d i r e c t l y attributed to the development of water desalination membranes, the invention of modified s i l i c o n e membranes and polysulfone membranes was more influenced by the extension of knowledge of transport, sorption and d i f f u s i o n of gases i n polymers (24-27). In p r i n c i p l e , rubbery polymers exhibit the highest gas permeabilities at the lowest s e l e c t i v i t i e s , and,



11.

FINKEN

Asymmetric Membranes for Gas Separations

251

unfortunately, they have a strong tendency to distend under higher pressure differences. To solve this problem of mechanical i n s t a b i l i t y , Ward and co-workers (28-30) pioneered i n 1976 the use of block copolymers f o r gas permeation membranes. They combined the high gas permeability of rubbery polymers l i k e s i l i c o n e with the function of glassy polymers l i k e polycarbonate which act as cross l i n k s between the rubbery segments, because glassy chain segments tend to aggregate i n r i g i d , microscopic domains. To increase the throughput, they t r i e d to decrease the membrane thickness (31) by using a method known i n b i o l o g i c a l membrane research as the Langmuir-Blodgett technique. This method i s used to b u i l d monomolecular layers of film-forming substances at water-air interfaces. A schematic of a Langmuir trough i s depicted i n Figure 1. The f i l m formation at the air-water interface i s accomplished by adding drops of d i l u t e film-forming solutions to the water surface or by allowing the solution to flow down evenly onto a p l a s t i c block i n s t a l l e d a few millimeters above the water surface. I f the f i l m forming agent has surface a c t i v i t y , the solution spreads spontan eously on the water surface. Amphiphilic molecules with a hydro phobic, l y o p h i l i c p a r a f f i n t a i l and a hydrophilic, polar headgroup best meet the requirements f o r film-forming materials. Polymers that do not spread spontaneously can, nevertheless, be readily adsorbed i n t e r f a c i a l l y from solution i n a v o l a t i l e , water-immiscible solvent l i k e hexane, cyclohexane, benzene, dimethylformamide, dimethylsulfoxide or their mixtures. The structure of the f i l m can be made dense by decreasing surface area. This i s achieved by moving the p l a s t i c block, which i s p a r t i a l l y immersed i n the water, toward the e x i s t i n g f i l m . A t y p i c a l surface-pressure/surface-area isotherm, along with s i m p l i f i e d structures of monolayers, i s given in Figure 2. The surface pressure, or the difference between the surface tensions of the solute and solvent, changes through a series of steps from a gas to a f l u i d and f i n a l l y to a s o l i d state when the f i l m i s compressed (32,33). Ward and co-workers adapted t h i s technique to the preparation of u l t r a t h i n silicone/polycarbonate membranes and advanced i t by spreading the d i l u t e solution f i r s t on a limited water surface between two p l a s t i c blocks and then moving the blocks apart. This method spreads the polymer solution continuously, reducing the^^ thickness of the f i l m . A nitrogen permeability rate of 8.6 χ 10 cm (STP)/cm -s-cmHg and an 0 /N s e l e c t i v i t y of 2.3 are reported for a 1000 £ t h i n f i l m . Applications under i n v e s t i g a t i o n f o r t h i s type of membrane are the production of oxygen-enriched a i r f o r i n d u s t r i a l combustion and b i o l o g i c a l degradation processes or the production of nitrogenenriched a i r with less than 9% oxygen to inert f u e l tank spaces and vent l i n e s of a i r c r a f t . A portable oxygen-enricher i s commercially available from Oxygen Enrichment Co. (Schenectady, NY) f o r r e s p i r atory disease oxygen therapy. The l a t e s t major innovation, and also economically the most successful breakthrough, was achieved by Henis and T r i p o d i (34-38) with t h e i r invention of the resistance model (RM) f i b e r membranes i n l a t e 1979. In agreement with the h i s t o r y of reverse osmosis membrane development, they selected a high-strength, glassy polymer as membrane material, such as polysulfone with a glass t r a n s i t i o n 9

1

1




I Figure 1.

Air£ Water

P r i n c i p l e s of a Langmuir trough

solid

',

Λ

-

\

fluid

^

\ goseoLB"— //-J

—

30

1

Figure 2.

//-n

1

50 Area

-=v--_r---

100 (λ /molecule) 2

Typical surface pressure/area isotherm


-


11.

FINKEN


253

temperature of 180°C. The ^Intrinsic hydrogen permeability of polysulfone i s 1.2 χ 10~ cm (STP)-cm/cm -s-cmHg and i t s s e l e c t i v i t y for the system hydrogen/carbon monoxide i s 40. Compared to s i l i c o n e with a value of 5.2 χ 10" cm (STP)-cm/cm -s-cmHg, the hydrogen permeation through polysulfone i s lower by a factor of 43.3. But t h i s disadvantage i s more than compensated for by the a b i l i t y to form polysulfone hollow f i b e r s with a great volumeto-surface r a t i o , a technology that cannot be used for the rather soft rubbery polymers l i k e s i l i c o n e . In addition, the hydrogen/ carbon monoxide s e l e c t i v i t y for s i l i c o n e i s only 2.1 for s i l i c o n e . In contrast, c e l l u l o s e acetate shows a s e l e c t i v i t y for the s^me gas pa^r of 46.3 ^and i t s hydrogen permeability i s 5.0 χ 10 cm (STP)-cm/cm -s-cmHg. Therefore, i t s transport behavior i s similar to that of polysulfone which seems to be not surprising, since the glass t r a n s i t i o n of c e l l u l o s e acetate i s i n the same temperature range of 180 to 185°C l i k e polysulfone. A serious problem which Henis and T r i p o d i had to overcome was the production of reproducible hollow f i b e r s free of imperfections. In contrast to reverse osmosis, skin layer fine pores with diameters larger than those of the penetrant gas molecules e f f e c t a viscous or pore flow through the membranes, which reduces d r a s t i c a l l y t h e i r separation c a p a b i l i t i e s . This i s because the gas v i s c o s i t i e s and gas d i f f u s i o n c o e f f i c i e n t s are several orders of magnitudes lower and higher, respectively, than those of l i q u i d solutions and solutes. From an engineering point of view, the higher d i f f u s i o n c o e f f i c i e n t of gas i s advantageous, since the phenomena of concen t r a t i o n p o l a r i z a t i o n on the upstream side of the membrane can be neglected, which allows a s i m p l i f i c a t i o n of the module design. Whereas the surface porosity i s d i f f i c u l t to prevent i n flat-sheet membrane technology, i t i s much more vexing i n hollow-fiber technology. In addition to the two concomitant requirements of very t h i n and defect-free skin-layers of integral-asymmetric membranes, the porosity of the supporting sublayer must be optimized simultaneously to prevent i t s collapse when subjected to high pressures and to minimize i t s resistance to gas flow. Therefore, Henis and T r i p o d i separated the problem of surface porosity from the actual preparation of the polysulfone f i b e r s by vacuum coating a t h i n layer of s i l i c o n e rubber inside the f i b e r bores. A schematic view of the cross-section of an RM hollow-fiber and the e l e c t r i c a l c i r c u i t analog i s presented i n Figure 3. In e f f e c t , the s i l i c o n e i s sucked into the pores and f i l l s them. The t o t a l flow resistance can be considered analogous to an e l e c t r i c a l c i r c u i t consisting of four resistances i n s t a l l e d i n a s e r i e s p a r a l l e l configuration. A gas must f i r s t pass through the coating layer (R^), then either through the pores f i l l e d with coating material (R ) or through the skin-layer of the f i b e r (R^), and f i n a l l y through the porous sublayer (R^) to the f i b e r bores. The t o t a l resistance i s given by 2

1 R

l

+ R,

+

ï 2 R

+

R R

3

Because s i l i c o n e rubber i s a highly permeable polymer, the


gas



254

transport through the underlying less permeable, glassy skin-layer of polysulfone i s expected to be influenced only s l i g h t l y by the coating-layer. This i s v e r i f i e d i n Table I I I where the resistances to permeate gas flows for the d i f f e r e n t parts of RM f i b e r s tjave been calculated with the assumption of a membrane area of 1 m . On the basis of i n t r i n s i c permeabilities (36), the c a l c u l a t i o n reveals the lowest flow resistances for the coated s i l i c o n e layer. Compared to t h i s , the flow resistance of the polysulfone skin-layer to hydrogen i s higher by a factor of 4.4, and resistance to carbon monoxide i s higher by a factor of 84. The greatest resistances to the permeating flow are created by the pores f i l l e d with s i l i c o n e . Their values are 4 to 5 orders of magnitude higher than the other resistances. With the r a t i o n a l assumption of n e g l i g i b l e resistance to flow offered by the porous sublayer and with the result that the pore resistance i s much greater than the skin-layer resistance, the r e l a t i o n f o r the t o t a l resistance s i m p l i f i e s to R

t

= R

1

+

R

3

.

According to Table I I I , the contribution of the s i l i c o n e layer to the flow of the less permeable carbon monoxide i s only 1.2% and can be neglected, but i t does contribute remarkably to the flow of the more permeable hydrogen by approximately 20%. Correspondingly, the H^/CO separation factor of the selective polysulfone skin-layer i s reduced by 20% from 40 to 33, because the i d e a l separation factor i s by d e f i n i t i o n the r a t i o of the permeability rates or - v i c e versa - the r e c i p r o c a l of the r a t i o of the flow resistances. With t h i s invention, Henis and T r i p o d i discovered a simple and ingenious solution to the problem of surface porosity. They found a l i m i t i n g value f o r the surface porosity. I f i t exceeds the value of 10 , the s e l e c t i v i t y of the f i b e r i s destroyed almost completely (36). The RM hollow-fibers are employed i n the PRISM separators of Monsanto (St. Louis, MO) which have been successfully tested since 1979 i n more than 40 i n d u s t r i a l plants i n the chemical and refinery field. They are being used f o r the recovery of hydrogen from waste-gases and purge-streams that contain varying amounts of N^, Ar, CO, C 0 C^ to C^ p a r a f f i n s , and C^ to Cg aromatics. 2>

The Influence of the Physical Structure of Integral-asymmetric Blend Membranes on Carbon Dioxide/Methane Separation

CA

After the 1982 decision by the GKSS Research Center (a national laboratory i n Geesthacht, Federal Republic of Germany) to s h i f t the research and development a c t i v i t i e s i n the f i e l d of membrane technology from seawater desalination to selective permeation for gas separation, several commercial RO membranes were f i r s t i n v e s t i gated f o r t h e i r u t i l i z a t i o n i n gas separation by determination of t h e i r gas permeability rates. The permeation apparatus designed for a constant-pressure/variable-volume method i s depicted i n Figure The flow test unit with an e f f e c t i v e membrane area of 35.26 cm i s supplied with either a pure gas or a gas mixture from gas cylinders. The upstream gas pressure i s adjusted by a pressure regulator and monitored by a precision manometer, while the


11.

FINKEN

255


Table I I I :

Resistances

to H

2

and CO flows through RM hollow f i b e r s

made of silicone-coated polysulfone

Fiber part

Resistance


H

s i l i c o n e layer ( R )

3 (s-cmHg/cm ) to CO

2

0.19

x

pores f i l l e d with silicone ( R )

0.40

, 1.00 x 10*

2

_ 1.75 χ 10

polysulfone skin-layer

0.83

33.35

t o t a l f i b e r resistance

1.03

33.75

( ν

Note:

I t was assumed that area 6

= 1 m^; 6

10" ;

thickness λ. = 1 χ 1θ" m; and i

= A^; A^/A = 1.9 x 2

7

- £ - l χ 1θ" m; 3 2

and that the i n t r i n s i c permeabilities i n cm (STP)-cm/cm -s-cmHg were as follows (36): gas H

2

CO

silicone

polysulfone

5.2 χ 10"

8

1.2 χ 10~

9

2.5 χ 10"

8

3.0 χ 1 0 ~

U


256

M A T E R I A L S SCIENCE O F S Y N T H E T I C M E M B R A N E S


RM

gas

hollow fiber

Electrical

separation

gas

—$

circuit

analog

analysis

—

-10

12

8

4·«·

Figure 4.

•13

2

Diagram of the permeation apparatus (Inflow test unit; 2=thermostated water bath; 3=gas cylinder; 4=pressure regulator;

5=pressure

gauge;

7,8=^lowmeters; 9-four-way

6=back-pressure

valve;

valve; 10=to gas chromato-

graph; ll=absorption column; 12=flowmeter;

13=vent l i n e )



11.

FINKEN


257

downstream pressure i s equivalent to the atmospheric pressure. The temperature i s controlled by placing the test unit i n a thermostated water bath. The flow of the retentate i s regulated by a back pressure valve. Permeate and retentate flows are measured, at ambient pressure and temperature, by water or soap-bubble displace ment i n graduated glass cylinders or burettes. Gas mixtures are analyzed by gas chromatography or absorption techniques. The permeability rates of the commercial RO membranes for helium, oxygen and carbon dioxide along with the i d e a l separation factors for helium/nitrogen, nitrogen/oxygen and carbon dioxide/ methane are recorded i n Table IV. The data were determined with pure gases at a temperature of 25°C and at an upstream pressure of 40 bar. The readings were taken after the system had reached steady-state conditions. The permeation experiments with different gases were carried out i n the following sequence: helium, oxygen, nitrogen, methane and carbon dioxide. In general, the commercial RO membranes had high permeability rates at low s e l e c t i v i t i e s . As the s e l e c t i v i t i e s did not approach the i n t r i n s i c values of the polymers, i t was concluded that the selective skin or composite layers were not free of pores and imperfections. One type of composite membrane (UOP, RC 100) could not be investigated, because the composite layer was so b r i t t l e that i t tore when assembled i n the test u n i t . These r e s u l t s demonstrate that RO membranes do not have the c a p a b i l i t y to separate gases. There i s one remarkable exception: the CA blend membranes showed a high s e l e c t i v i t y of 41.8 f o r He/N^ at a considerable permeability rate. The prospective gas separation properties of CA blend membranes for carbon dioxide and methane are also outlined i n Table V where some l i t e r a t u r e data are summarized. This table shows again that glassy polymers with a r i g i d back-bone and, hence, a high glass t r a n s i t i o n temperature, such as polyimide, have the highest selec t i v i t y f o r CO^/CH, of 65.0, and that rubbery polymers, such as dimethy1silicone,rhave thejiighest carbon dioxide permeability rate of 1.08 χ 10" cm (STP)/cm -s-cmHg. While the separation c a p a b i l i t y of s i l i c o n e i s 19 times lower than that f o r polyimide, i t s permea t i o n rate f o r carbon dioxide i s 13,500 times higher. The transport properties of c e l l u l o s e acetate are situated between these two polymers: t h e i r s e l e c t i v i t y i s ten times higher than s i l i c o n e , and two times lower than polyimide, whereas t h e i r permeability rate i s 32 times higher than polyimide and 420 times lower than s i l i c o n e . It may be also interpreted from the data of Table V that the gas separation properties of CA films are almost independent on t h e i r degree of substitution. This i s inconsistent with the fact that the glass t r a n s i t i o n temperature strongly depends on the hydroxyl content of the c e l l u l o s e acetate polymers. With the loss of hydrogen bonding interactions, the glass t r a n s i t i o n drops from 185°C f o r a c e l l u l o s e diacetate (d.s.=2.45) to 114°C for a c e l l u lose t r i a c e t a t e (d.s.=2.86). There are two reasons which suggest that the CO^ permeabilities of c e l l u l o s e t r i a c e t a t e should be higher than those of c e l l u l o s e diacetate: the CO^ sorption i s supposed to increase with the degree of acetylation due to favorable e l e c t r o s t a t i c interaction between the dipoles of the carbonyl groups of the ester and the polar carbon dioxide molecules. The C0 d i f f u s i o n should be eased by the lack of polymer-polymer i n t e r ?



Manufacturer

Envirogenics

CA blend

Nitto

polyurethane

146 1292

CO.

2

41.8

2.9

0.37

1681

13,962

6542

-

-

5811

-

-

5080

19,001

1.8

1.6

3.1

0

1.6

0.3

V°2

0.6

9.0

4.2

2

4

2)

«

C0 /CH

Ideal Separation Factor He/N

tore apart, when assembled i n test unit

1828

1389

He

Permeability Rate 1)

-7 3 2 3 2 1) i n 10 cm (STP)/cm -s-cmHg; to obtain m (STP)/m -s-MPa multiply by 7.501 2) for C0 /N

2

FilmTec

Nitto

ρ οlyethy1eneimine

Osmonics

unknown (RO 98)

polyamide (FT 30)

UOP

polyetherurea

t h i n - f i l m composite membranes

Hydranautics

c e l l u l o s e acetate

integral-asymmetric membranes

Membrane polymer

Table IV: Pure gas permeability rates of commercial reverse osmosis membranes at 25°C and 40 bar


I

m

2 m 2 w >

Π

Η Χ

Ο τι

η m

m

Q

£

m

25

2

11.

FINKEN


Table V:

Cellulose acetate i n comparison to other polymers

Τ


1 }

P/A(C0 )

2

)

3

a*

2

g

Polymer

259

)

C0 -CH^, 2

symmetric membranes

Reference

^

PI

400

0.08

65.0

(42,43)

CA

185

3.20

34.9

(20)

CA/CTA

170

CTA

155

2.28

33.5

(43)

-117

1080.00

3.4

(44)

1610.00

26.0

(20)

Silicone

5

)

2.24

34.9

(20)

asymmetric membranes CA/CTA

170

1) glass t r a n s i t i o n temperature T^ i n °C -

-7

2) permeability rate P/i, i n 10 3

3

2

cm (STP)/cm -s-cmHg; to obtain

2

m (STP)/m -s-MPa multiply by 7.501 3) i d e a l separation factor a* f o r the gases carbon dioxide and methane 4) with a f i l m thickness of 25 um 5) estimated


M A T E R I A L S SCIENCE OF SYNTHETIC M E M B R A N E S


260

actions. V e r i f i c a t i o n of t h i s hypothesis i s currently under investigation. The most important requirements of high s e l e c t i v i t y and high permeability for the more permeable CO gas seem to be met by the asymmetric CA blend membranes. They exhibit the permeability rates of rubbery materials and the s e l e c t i v i t i e s of a glassy polymer with an intermediate high glass t r a n s i t i o n temperature. These f i r s t r e s u l t s and suggestions stimulated further i n v e s t i gations of CA blend membranes for gas separation. At a given chemical composition of the CA blend, that i s at a r a t i o of c e l l u lose diacetate to t r i a c e t a t e of 1:1, and at a given phase inversion process for asymmetric membranes which i s described i n d e t a i l e l s e where (39-41), d i f f e r e n t kinds of membranes could be prepared by applying various post-treatment methods, such as annealing, freezedrying and solvent-drying. As mentioned above, gas separation membranes should be i n an u n p l a s t i c i z e d , s o l i d state to maintain t h e i r separation properties over a long period of time. Therefore, the gel membranes swollen i n water must be transferred to a dry, s o l i d state. Freeze-drying. To improve the CA blend membranes for gas separa t i o n , attempts were made to freeze-dry the membranes. Normally, the process of freeze-drying i s applied to b i o l o g i c a l substances which are susceptible to s t r u c t u r a l damages and to loss of proper t i e s . I t i s the preferred method to pretreat s e n s i t i v e specimens for investigation by scanning electron microscopy. The drying of wet membrane samples i s s i m i l a r l y aimed at the maintenance of t h e i r asymmetric structure. The microporous sublayer with i t s high water content i s p a r t i c u l a r l y susceptible to collapse during the drying process and thus to enhance the thickness of the skin-layer. As the c r y s t a l size of the frozen water favorably decreases with increasing rate of cooling, the wet membranes were f i r s t shockfrozen at the nitrogen melting point. By this means, i c e c r y s t a l growth during the phase t r a n s i t i o n , which would cause volume expansion and damage to the membrane structure, was almost prevented. The vitreous i c e was then allowed to sublimate under vacuum. The temperature during sublimation influenced the rate of gas permeation (Figure 5), which was measured with pure gases at the test conditions of 25°C and 40 bar. The £0 rate increases from 1.85 x 10 to 2.40 χ 10 cm (STP)/cm -s-cmHg with an increase i n the sublimation temperature from -100°C to -40°C. The corresponding rates for helium, nitrogen and methane are reduced by 76, 91 and 91%, respectively, with the same 60°C increase i n sublimation temperature. The plot of the permeability r a t i o s for the system CO^/CH, and He/N i n dependence of sublimation tempera ture (Figure o) snows that the He/N s e l e c t i v i t y i s almost not affected by the temperature of sublimation, but that the C0«/CH^ s e l e c t i v i t y exponentially increases to values above 100 with increasing sublimation temperature. This effect of sublimation temperature on the physical structure of the membrane has not yet been investigated by independent methods l i k e scanning electron microscopy. But the r e s u l t s of these permeation measurements may be interpreted i n terms of a s o l u t i o n - d i f f u s i o n model which was f i r s t introduced to describe the transport behavior through reverse ?

2

9


11. FIN Κ EN


261

P/l m /m hbar 3

2

7.

6.


5.

i*

3

2

1

180

Figure 5.

200

220

2^0

K

Permeability rate, P A , as a function of sublimation temperature, T, during the freeze-drying of CA blend membranes 3

(pure

gases, 25°C,

40 bar; to obtain

2

4

cm (STP)/cm -s-cmHg multiply by 3.69 x 1 0 ~ )

α

Figure 6.

Ideal separation factor, a, as a function of sublimation temperature, T, during the freeze-drying of CA blend membranes (pure gases, 25°C, 40 bar)



262

M A T E R I A L S SCIENCE OF S Y N T H E T I C M E M B R A N E S

osmosis membranes. In contrast to l i q u i d - s o l u t e separations, where the extremely low s o l u b i l i t y of s a l t s i n the membrane excludes them almost completely from transport through membranes, the d i f f u s i v i t y term i n gas-gas separations mainly contributes to the o v e r a l l permeations of permanent gases through glassy polymers. Exceptions from t h i s general rule occur only with those penetrating gas molecules which strongly i n t e r a c t with the polymer. In t h i s case, the s o l u b i l i t y may predominate the d i f f u s i v i t y . As described above, the main problem i n freeze-drying of asymmetric membranes i s the prevention of ice c r y s t a l growth during the very short period of i n i t i a l shock-freezing. But t h i s may also hold true for the long period of time of actual freeze-drying, during which the i c e inside the membrane must sublimate to water vapor molecules and then d i f f u s e to the membrane surface. At the same time, the countercurrent effect of c r y s t a l growth may take place, which i s faster the higher the sublimation temperature i s . By t h i s , the sponge-like structure of the supporting layer may be p a r t l y damaged and compacted when pressurized. The e f f e c t i v e thickness of the s e l e c t i v e b a r r i e r layer w i l l , i n turn, grow and hence the permeation rate w i l l decrease with increasing sublimation temperature as shown for helium, nitrogen and methane (Figure 5). The reduction i n d i f f u s i o n rate should also apply for CO^. Therefore, the observed increase i n CO^ permeation can only be explained by an overproportional enhancement i n CO^ s o l u b i l i t y , which i s the dominant factor for the permeation because of strong interactions between CO- and c e l l u l o s e acetate. It i s clear from the "dual-mode sorption* model that the s o l u b i l i t y depends on the nature of the polymer and penetrating gas and on the number and size of pre-existing holes i n the polymer. Only the l a t t e r can be influenced by the freeze-drying process, and i t i s evident that the number of pre-existing holes which can be f i l l e d with CO^ should be increased i n the membranes prepared at higher sublimation temperatures. A comparison of gas transport rates of freeze-dried CA blend membranes with those treated by simple air-drying shows that freeze-drying produces membranes with CO^/CH^ s e l e c t i v i t i e s higher by a factor of 27 (see Table VL). Jhe CdL permeation i s faster and reaches a value of 239.4 χ 10* cm (STP)/cm -s-cmHg. However, the s e l e c t i v i t i e s f o r smaller penetrants such as He/N are reduced by a factor of 3.2. The reason for the s i g n i f i c a n t a l t e r a t i o n s of gas transport with varying post-treatment procedures i s not yet well understood. The r i s k of damaging the s e l e c t i v e skin-layers i s increased when membranes are dried i n an uncontrollable and higher rate at ambient conditions. Whereas the low s e l e c t i v i t y of CO /CH, for a i r - d r i e d membranes supports t h i s opinion, the high He/ΙΓ s e l e c t i v i t y does not. Annealing. The f i n a l properties of Loeb-Sourirajan type CA membranes are created by annealing the membranes i n hot water at temperatures between 70 and 100°C. In e f f e c t , s a l t r e j e c t i o n increases at the expense of permeate water flux with increasing temperature. Two mechanisms may be discussed to explain t h i s



264

M A T E R I A L S SCIENCE O F S Y N T H E T I C M E M B R A N E S

behavior. F i r s t , the active skin layer may either shrink, thereby reducing i t s thickness but enhancing i t s density, or the skin-layer may be thickened by p a r t i a l conversion of the adjacent, microporous sublayer into the dense skin layer. Secondly, the enhanced mobility of the CA polymer segments during annealing may cause i n t r a molecular or intermolecular hydrogen bonds to change to more favor able, exothermic interactions leading to a higher degree of polymer c r y s t a l l i n i t y . More c r y s t a l l i n e domains generally reduce sorption and also represent b a r r i e r s to d i f f u s i o n . The effect of wet annealing on gas permeation of CA blend membranes i s demonstrated i n Figure 7. The wet membranes were immersed i n a hot water bath at temperatures of 90, 95 and 100°C for 15 minutes, quenched to ambient temperature, soaked i n ethano^ and a i r - d r i e d . - Theu permeability rates of He and Ν of 11.0 χ 10 and 1.2 χ 10" cm (STP)/cm -s-cmHg were not affected by this_^ procedure. -The same i s true f o r carbon dioxide with 6.3 x 10 cm (STP)/cm -s-cmHg up to 90°C. But above t h i s temperature, the CO permeation grows rapidly by 185% from 90 to 100°C. At 100°C, the permeability permeability rate of 1.8 χ 10 cm (STP)/cm -s-cmHg (see Table VI). Compared to freeze-drying, the annealed membranes show permeability rates f o r helium and carbon dioxide lowered by 75 and 25%, respectively. The i d e a l separation factor for CO /CH^ i s , however, increased by 28%. The observed constant permeability rates for helium and nitrogen indicate that the f i r s t of the two above proposed effects of annealing does not hold since changes i n thick ness or density of the active layer would cause a direct response i n permeation. Only i n the rare case that an increased density i s exactly compensated f o r by a decreased thickness would the permea t i o n be kept constant. The increase i n CO^ permeation, however, may be explained s a t i s f a c t o r i l y by the second mechanism; that i s , by configurational changes of polymer interaction due to heat treatment of the CA polymer p l a s t i c i z e d by water. The steep increase above 95°C indicates that second-order transitions combined with a more randomly distributed order of the polymer can take place. As t r a n s i t i o n temperatures are generally depressed by p l a s t i c i z a t i o n , i t i s not surprising that this wet t r a n s i t i o n temperature i s somewhat lower than those observed f o r dry, u n p l a s t i cized polymers. The glass t r a n s i t i o n of dry, s o l i d c e l l u l o s e diacetate and t r i a c e t a t e are i n the range 180 to 190°C and i n the range 150 to 160°C , respectively, depending on hydroxyl content, because the enhanced acetyl-content lowers the hydrogen-bonding i n t e r a c t i o n between the polymer chains. Moreover, four second-order transitions are reported f o r c e l l u l o s e diacetate at temperatures of 15, 40 to 60, 70 to 90 and 114 to 120°C, and two are described for c e l l u l o s e t r i a c e t a t e at temperatures of 30 to 46 and 105 to 120°C (45). Therefore, the transport of CO- through CA blend membranes may be changed when annealed at 100° (Γ which i s above the glasst r a n s i t i o n temperature of wet CA blend polymers. The exponential permeability increase of C0~ may be attributed to the second-order transitions described above leading to a higher s o l u b i l i t y of CCL i n c e l l u l o s e acetate and a stronger p l a s t i c i z a t i o n at elevated CO^ pressures.


11.

FINKEN


265

Solvent-drying. After annealing at 100°C, the absorbed water inside the membranes was removed by a proprietary procedure. I t s effect on gas permeation i s given i n Table VI. No effect i s observed on the helium and carbon dioxide permeability rates. But the permeab i l i t y rates of nitrogen and methane are considerably lowered so that the separations of both systems, helium/nitrogen and carbon dioxide/me thane, are enhanced. With a value of 44.1 for helium/ nitrogen, the i d e a l separation factor of the external reference membrane i s s l i g h t l y exceeded, whereas the separation factor for C0 /CK^ of more than 2500 i s improved by a factor of 600. It i s evident that the permeability rates of gases s i m i l a r i n size and shape l i k e nitrogen and methane depend on the actual state of the polymers. The transformation of gelled polymers to solids by the drying procedure may change t h e i r state i n such a manner that the polymer density increases and the "free volume" decreases r e s u l t i n g i n lowered permeability rates. Exceptions from t h i s general rule apply only f o r molecules such as helium which are very small and carbon dioxide which are able to p l a s t i c i z e the CA polymers.


2

The Influence of Temperature and Properties of CA Blend Membranes

Pressure on

the Transport

Temperature. The outstanding transport behavior of the carbon dioxide/cellulose acetate blend system was again experienced when the test conditions were changed. The temperatures of the thermostated water bath was set between 18 and 50°C, while the upstream pressure of the pure gas was kept constant at 40 bar. As depicted i n an Arrhenius plot of logarithmic permeabilities versus r e c i p r o c a l temperatures i n Figure 8, the permeations of He, N and CH, were accelerated by increasing temperatures, and the CO^ permeation was slowed down. Consequently, the He/N separation i s almost independent of temperature as shown i n Figure 9 whereas the CO^/CH^ separation declines steeply from 2500 to 500. I t i s a matter of fact that the s o l u b i l i t y of a given gas i n polymers generally decreases with increasing temperature while i t s d i f f u s i v i t y increases. In cases where the permeability of gases through polymer films increases with temperature, such as depicted i n Figure 8 f o r He, N and CH^,, i t i s clear that the d i f f u s i v i t y i s the dominant factor f o r permeation. In those cases where the gas permeability decreases with temperature, i t may be concluded that the s o l u b i l i t y term mainly influences permeation. A reduction of the CO- permeability by 71% i s calculated f o r CA blend membranes from Figure 8 where the temperature raises from 20 to 60°C. Comparison to the temperature-dependence of s o l u b i l i t i e s f o r lowmolecular-weight analogs confirms t h i s behavior. For the same temperature increase from 20 to 60°C, the s o l u b i l i t i e s of C0 i n water and acetone decline by 63 and 95%, respectively (46). Thus, the s o l u b i l i t y decrease i s i n the same order of magnitude or even higher which compensates f o r the enhanced d i f f u s i v i t y at higher temperatures. 2

2

2

?

Pressure. Experiments with variable pressures were conducted with pure gases at a constant temperature of 25°C. Pressures between 10


M A T E R I A L S SCIENCE O F SYNTHETIC

266

MEMBRANES

P/l m /m hbar 3

2


5

, He

-y ? N

90

Figure 7.

95

100

» δ

τ

°C

Gas permeability rate, P/Jl, as a function of annealing temperature, 6 ,of asymmetric CA blend membranes 3 2 T

gases, 25°C, 40 bar; to obtain

(pure

cm (STP)/cm -s-cmHg

4

multiply by 3.69 χ 10*~ )

1

Figure 8.

A

Temperature-dependence

of

permeability

rate

of

asymmetric CA blend membranes (pure gases, 40 bar, Ρ 3

1 m (STP)/

2

m -h-bar,

to obtain

3

2

cm (STP)/cm -s-cmHg

4

multiply by 3.69 χ 1θ" )


=


11.

FINKEN

Figure 9.


Temperature-dependence of i d e a l separation factor, asymmetric CA blend membranes at 40 bar.


268


and 80 bar were set for the gases He, Ν and CH^, and between 10 and 40 bar for CO^. Semi-logarithmic plots of permeability rates versus applied upstream pressure are presented i n Figure 10. The He permeation i s pressure-independent, but the permeations of the three other gases increase with pressure. Again, the steep slope of CO^ i s unique and w i l l be discussed i n d e t a i l . Due to these f a c t s , the i d e a l separation factor f o r C0 /CH, increases more than proportionally and that for He/N- decreases l i n e a r l y with increasing pressure (Figure 11). The unique behavior of c e l l u l o s e acetate was supported recently by Koros (42). He found that of a l l glassy polymers investigated, c e l l u l o s e acetate had the highest s o l u b i l i t y s e l e c t i v i t y , defined as the r a t i o of the s o l u b i l i t i e s of two gases C0 and CH^. Its extremely high s o l u b i l i t y value of 8.9 measured at 35°C and 20 atm exceeds the values of the other polymers l i s t e d below by factors of 2.5 to 4.0. Along with the mobility s e l e c t i v i t i e s which were defined as the r a t i o of the d i f f u s i o n c o e f f i c i e n t s of two gases i n the polymer, Koros predicted that the CO- permeabilities for c e l l u l o s e acetate should increase by 138% i r the pressure increases from zero to 20 atm. The CO^ permeabilities of a l l the other glassy polymers: PPO, PS, PC, polysulfone and polyimide were found to decrease by 30 to 35% with pressure. By extrapolating^the^CO- curve of2Figure 10 to zero pressure, a value of 17.5 χ 10~ cm (STP^/ cm^-s-cmHg £s obtained. Compared to the value of 42.6 χ 10 cm (STP)/cm -s-cmHg at 20 bar an increase by 143% i s calculated. This experimental finding v e r i f i e s the above mentioned prediction. The extreme sorption of carbon dioxide i n c e l l u l o s e acetate causes a strong p l a s t i c i z a t i o n of the polymer at elevated pressures. By t h i s e f f e c t , CO^ shows the highest permeability rates of a l l permanent gases investigated. Further investigations with gas mixtures containing carbon dioxide w i l l demonstrate how the permeability of the less soluble and less permeable methane i n the gas mixture i s increased by the CO- p l a s t i c i z a t i o n . It i s clear that, furthermore, the i d e a l separation factor given i n t h i s paper w i l l be reduced to more r e a l i s t i c values which, when measured i n the gas mixtures at zero recovery, w i l l not exceed the permeability r a t i o calculated by extrapolating the permeability-pressure plots to zero pressure. An upper l i m i t of about 300 i s computed f o r the CO^/CH^ system from Figure 10 for zero pressure which i s ten times higher than that derived by Koros (47).


9

9

Conclusions and Recommendations The history of the membrane developments for reverse osmosis and gas permeation shows that because of inherent differences, i t i s not possible to simply apply the techniques and materials from one separation technology to the other. The success of the resistancemodel hollow-fiber technology which i s based on the glassy-fiber technology invented f o r reverse osmosis, demonstrates the necessity to search f o r advanced techniques to prepare more selective membranes free of imperfections, rather than to look for new, unavailable materials. The interfacial-polymerization technology f o r the preparation of u l t r a - t h i n composite membranes with i t s s e l f - s e a l i n g c a p a b i l i t y


11.

FINKEN

269


2 • 1 .


-1 -2 -3 -U

-5 -6

Figure 10. Pressure-dependence of permeability rate of asymmetric CA blend membranes (pure gases 25°C, Ρ* Ξ 3 2 3 2 m (STP)/m -h-bar; to obtain cm (STP)/cm -s-cmHg multiply 4

by 3.69 χ 10~ )

2500 C0 / CH^ 2

1500

500

U

36 He/N

2

20

UQ

60

80

bar

Figure 11. Pressure-dependence of i d e a l separation factor,α asymmetric CA blend membranes at 25°C


, of


270

seems to be one promising way to overcome the problem of surface porosity. Further approaches to meet the requirement of high s e l e c t i v i t y may include the blending of glassy and rubbery polymers, the chemical a l t e r a t i o n of the dense skin-layer of integral-asymmetric membranes and morphological variations of dense polymer films by proper post-treatment—as exemplified i n t h i s paper for CA blend membranes.


Acknowledgment s The author i s g r a t e f u l to Mr. H. P. Witt f o r h i s technical a s s i s tance i n preparing the membranes and i n performing the experimental program. The author also acknowledges Mrs. R. Ebermann f o r her help i n producing t h i s manuscript.

Literature Cited 1. 2. 3. 4. 5. 6. 7. 8. 9.

10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Reid, C. E.; Breton, E. S. J . Appl. Polym. Sci. 1959, 1, 133. Loeb, S. In "Synthetic Membranes"; Turbak, A. F., Ed.; ACS Symp. Ser., No. 153: Washington, D . C . , 1981. Loeb, S.; Sourirajan, S. Adv. Chem. Ser. 1962, 38, 117. Riley, R. L.; Lonsdale, H. K.; Lyons, C. R.; Merten, U. J . Appl. Polym. Sci. 1967, 11, 2143. Rozelle, L . T . ; Cadotte, J . E.; McClure, D. J. In "Membranes from Cellulose and Cellulose Derivatives"; Turbak, A. F., Ed.; Interscience Publishers: New York, 1970. King, W. M.; Cantor, P. Α.; Schoellenbach, L . W.; Cannon, C. R. In "Membranes from Cellulose and Cellulose Derivatives"; Turbak, A. F., Ed.; ibidem. King, W. M.; Hoernschemeyer, D. L.; Saltonstall, C. W. In "Reverse Osmosis Membrane Research"; Lonsdale, H. K. and Podall, H. E., Eds.; Plenum Press: New York, 1972. Richter, J . W.; Hoehn, H. H. U.S. Pat. 3,567,632; March 2, 1971. Rozelle, L . T . ; Cadotte, J . E.; Cobian, K. E.; Kopp, C. V. In "Reverse Osmosis and Synthetic Membranes"; Sourirajan, S., Ed.; National Research Council of Canada: Ottawa, 1977; Chapter 12. Mahon, H. I. U.S. Pat. 3,228,876; Jan. 1966. Boeddeker, K. W.; Wenzlaff, A. Ger. Pat. 2,820,265 (1980). Riley, R. L.; Fox, R. L.; Lyons, C. R.; Milstead, C. E . ; Seroy, W. M.; Tagami, M. Desalination 1976, 19, 113. Goldsmith, R. L.; Wechsler, Β. Α . ; Hara, S.; Mori, K . ; Taketani, Y. Desalination 1977, 22, 311. Pusch, W.; Riley, R. L . Desalination 1977, 22, 191. Kurihara, M.; Kanamaru, N . ; Harumiya, N . ; Yoshimura, Y . ; Hagiwara, S. Desalination 1980, 32, 13. Cadotte, J . E.; Petersen, R. J.; Larson, R. E.; Erickson, E. E. Desalination 1980, 32, 25. Gantzel, P. K . ; Merten, U. I&EC Proc. Des. Dev. 1970, 9, 331. Stern, S. Α.; Sen, S. K . ; Rao, A. K. J . Macromol. Sci. Phys., Β 1974, 10, 507. Vos, K. D.; Burris, F. O. I&EC Prod. Res. Dev. 1969, 8, 84.


11.

FINKEN

20. 21. 22. 23. 24. 25.


26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.


271

Schell, W. J. ACS Div. Fuel Chem., Preprints 1975, 20, 253. Schell, W. J.; Houston, C. D. U.S. Pat. 4,134,742, 1979; Hydrocarbon Processing 1982, Sept., 249. Coady, A. B.; Davis, J . A. CEP 1982, Oct., 44. Schell, W. J.; Houston, C. D. ACS Symp. Ser. 223, Washington, D.C., 1983. Stannett, V. T.; Koros, W. J.; Paul, D. R.; Lonsdale, H. K.; Baker, R. W. Adv. Polym. Sci. 1979, 32, 69. Cabasso, I. In "Encyclopedia of Chemical Technology"; Wiley-Interscience: New York, 1980; Vol. 12. Stern, S. Α.; Frisch, H. L. Ann. Rev. Mater. Sci. 1981, 11, 523. Matson, S. L.; Lopez, J.; Quinn, J . A. Chem. Eng. Sci. 1983, 38, 503. Browall, W. R. U.S. Pat. 4,156,597, 1977. Kimura, S. G.; Lavigne, R. G.; Browall, W. R. U.S. Pat. 4,132,824, 1979; U.S. Pat. 4,192,842, 1980. Ward, W. J., III U.S. Pat. 4,279,855, 1981. Ward, W. J., III; Browall, W. R.; Salemme, R. M. J . Membr. Sci. 1976, 1, 99. Gaines, G. L. "Insoluble Monolayers at Liquid-Gas Interfaces"; Interscience: New York 1966. Fendler, J . H. "Membrane Mimetic Chemistry"; Wiley -Interscience: New York, 1982. Henis, J.M.S.; Tripodi, M. K. U.S. Pat. 4,230,433, 1980. --- Sep. Sci. Techn. 1980, 15, 1059. --- J. Membr. Sci. 1981, 8, 233. --- Polym. Sci. Techn. 1982, 16, 75. --- Science 1983, 220, 11. FInken, H. Desalination 1983, 48, 207. --- Desalination 1983, 48, 235. --- I&EC Prod. Res. Dev. 1984, 23, 112. Koros, W. J. "Membrane-based Gas Separations"; paper presented at the conference Membrane Technology in the 1980's; Sunriver, Oregon, Sept. 1983. Ellig, D. L.; Althouse, J . B.; McCandless, F. P. J . Membr. Sci. 1980, 6, 259. Kimura, S. G.; Walmet, G. E. Sep. Sci. Techn. 1980, 15, 1115. Stern, S. Α.; DeMeringo, A. H. J . Polym. Sci.: Polym. Phys. Ed. 1978, 16, 735. Stephen, H.; Stephen, T. "Solubilities of Inorganic and Organic Compounds"; Pergamon Press: New York, 1963. Chern, R. T.; Koros, W. J.; Hopfenberg, H. B.; Stannett, V. T. In "Materials Science of Synthetic Membranes"; Lloyd, D. R., Ed.; American Chemical Society: Washington, D.C., 1985.

RECEIVED August 6, 1984


Asymmetric Membranes for Gas Separations - American Chemical

Recommend Documents